The Rapid Emergence of Musical Pitch Structure in Human Cortex

• Published in: The Journal of neuroscience Vol. 40; no. 10; pp. 2108 - 2118
• Main Authors: Sankaran, Narayan, Carlson, Thomas A, Thompson, William Forde
• Format: Journal Article
• Language: English
• Published: United States Society for Neuroscience 04.03.2020
• Subjects: Acoustic mapping; Frequency; Learning algorithms; Machine learning; Magnetoencephalography; Music; Pitch; Temporal cortex; Waveforms; neural decoding; pitch perception; music perception; magnetoencephalography; multivariate pattern analysis; representational dynamics
• ISSN: 0270-6474; 1529-2401
• DOI: 10.1523/JNEUROSCI.1399-19.2020

In tonal music, continuous acoustic waveforms are mapped onto discrete, hierarchically arranged, internal representations of pitch. To examine the neural dynamics underlying this transformation, we presented male and female human listeners with tones embedded within a Western tonal context while recording their cortical activity using magnetoencephalography. Machine learning classifiers were then trained to decode different tones from their underlying neural activation patterns at each peristimulus time sample, providing a dynamic measure of their dissimilarity in cortex. Comparing the time-varying dissimilarity between tones with the predictions of acoustic and perceptual models, we observed a temporal evolution in the brain's representational structure. Whereas initial dissimilarities mirrored their fundamental-frequency separation, dissimilarities beyond 200 ms reflected the perceptual status of each tone within the tonal hierarchy of Western music. These effects occurred regardless of stimulus regularities within the context or whether listeners were engaged in a task requiring explicit pitch analysis. Lastly, patterns of cortical activity that discriminated between tones became increasingly stable in time as the information coded by those patterns transitioned from low-to-high level properties. Current results reveal the dynamics with which the complex perceptual structure of Western tonal music emerges in cortex at the timescale of an individual tone. Little is understood about how the brain transforms an acoustic waveform into the complex perceptual structure of musical pitch. Applying neural decoding techniques to the cortical activity of human subjects engaged in music listening, we measured the dynamics of information processing in the brain on a moment-to-moment basis as subjects heard each tone. In the first 200 ms after onset, transient patterns of neural activity coded the fundamental frequency of tones. Subsequently, a period emerged during which more temporally stable activation patterns coded the perceptual status of each tone within the "tonal hierarchy" of Western music. Our results provide a crucial link between the complex perceptual structure of tonal music and the underlying neural dynamics from which it emerges.